Islam, N., and M. Devarakonda. "An Essential Design Pattern for Fault-Tolerant Distributed State Sharing." Communications of the ACM, 39, 10 (October 1996), pp. 65-74.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... but are typically given low priority at design time. The Recursive Control pattern separates the control and service providing aspects of a real time program allowing each to be defined and modified independently. 5.1. 6 Recoverable Distributor Pattern The Recoverable Distributor Pattern [ID96], developed by Nayeem Islam Murthy Devarakonda (IBM T. J. Watson Research Center) assists developers of fault tolerant, state sharing distributed programs. This pattern focuses on the issues of performance and fault tolerance in distributed systems, and its usefulness is demonstrated through ....

Islam, N., and M. Devarakonda. "An Essential Design Pattern for Fault-Tolerant Distributed State Sharing." Communications of the ACM, 39, 10 (October 1996), pp. 65-74.

....in aims and scope. They offer solutions ranging from high level strategies for organising software to low level implementation mechanisms. Recently, there has been a growing number of patterns which express known concurrent behaviour of interacting entities over a possibly distributed platform [1, 5, 9, 16]. The proposed design pattern is based on the well known Observer pattern where a server process is in charge of monitoring a real time environment and informing a set of clients processes (called the observers) when changes in the monitored values (e.g. sensor values) occur. In this context, we ....

N. Islam and M. Devarakonda. An essential design pattern for fault-tolerant distributed state sharing. Communications of the ACM, 39(10):65--74, October 1995.

....The two most common approaches are N version programming and Recovery Blocks. A pattern implementing a generic version of these two approaches is given in [Dan97] An earlier version of this pattern, presented in pseudocode is shown in [Anc90] A related pattern is the Recoverable Distributor [Isl96], a pattern for fault tolerant, state sharing distributed processes, which also considers performance aspects. Exceptions Exceptions are a useful way to handle abnormal or unusual cases and contribute to the robustness of a system. Most advanced languages, including C and Java include a ....

N. Islam and M. Devarakonda, "An essential design pattern for fault-tolerant distributed state sharing", Comm. of the ACM, October 1996, 65-74.

....scope of the present design pattern approach. Thus, the proposal arose to connect design patterns closely with formal process specifications [15] Further directions of distributed design pattern development consider the enhanced performance and reliability requirements of distributed applications [8]. Our approach extends the descriptions of behavioural design patterns by formal process specifications. In general, this results in precise descriptions of dynamic aspects and correctness questions can be answered by means of formal methods. In particular, we use the specification technique cTLA ....

....of design patterns. We concentrate on the practical aspects of this approach and describe its application to an example system consisting of a central directory and local caches. The example is refined by the integration of the distributed observer pattern. This pattern has been introduced in [8] and is frequently used in the development of distributed applications. It is a refinement of the well known fundamental observer pattern [3, pgs. 293 ff] After describing the example, the distributed observer design pattern, and the formal specifications of both, we outline the formal ....

[Article contains additional citation context not shown here]

Islam, N., Devarakonda, M.: An Essential Design Pattern for Fault-Tolerant Distributed State Sharing. Communications of the ACM (39)10, 65--74.

....after a string of failed transactions. 6.2 Related Work Recent work in design patterns contains much in common with the process of defining and reusing the generic model described in this paper. Both approaches emphasize the specification of the core of a solution to a recurring problem [10]. However, the use of the generic model differs in two significant ways from the use of design patterns. First, the reuse of the generic model provided a mechanism for requirements validation rather than a design mechanism. The formally specified and validated components of the generic model were ....

N. Islam and M. Devarakonda, "An Essential Design Pattern for Fault-Tolerant Distributed State Sharing," CACM, Special Issue on Software Patterns, 39(10): pp. 65-71, October, 1996.

....objects contain, for copy policies, the Recoverable Object state. In this case the Memento pattern can be applied between the Recoverable Object and the Recovery Point where the former plays the role of Originator and the latter plays the role of Caretaker. The Recoverable Distributor Pattern [12], as the Customizable Object Recovery pattern, emphasizes the separation of concerns. It decouples failure detection from recovery. The Decorator Pattern [1] is used to add recovery responsibilities. The Recoverable Object corresponds to Concrete Component while Recovery Interface corresponds to ....

Nayeem Islam and Murty Devarakond. An Essential Design Pattern for Fault-Tolerant Distributed State Sharing. Communications of the ACM, 39(10):65--74, October 1996.

....utilization to DLS. 2.3 Scheduler Extensibility Our resource management system is extensible, in that it supports multiple scheduling strategies and permits the easy implementation and incorporation of new refined scheduling strategies. Octopus is designed using object oriented frameworks [22, 2, 16, 3, 17, 46]. This makes it possible, for example, to easily customize the DLS(s) to the needs of each computing domain. Similarly, the various aspects of PLSs can be customized to meet the needs of different types of partitions that concurrently exist in the system. The launcher DLS interface is also ....

N. Islam and M. Devarakonda. An Essential Design Pattern for Fault-Tolerant Distributed State Sharing. In Communication of the ACM, Oct. 1996.

....efficiency and partition utilization; in contrast, the input to the re allocation function of the DPS (notshown) of the PLS managing the space shared partitions is just the application efficiency. Scheduler Extensibility Octopus is an extensible system, designed using object oriented frameworks [8, 7, 1, 9], that supports multiple scheduling paradigms and permits the easy implementation and incorporation of new paradigms. Partitions with new scheduling policies are created by specializing two independent interactions: between a parent PLS and a child PLS and between a PLS and its NLSs. 4 One ....

N. Islam and M. Devarakonda. An Essential Design Pattern for Fault-Tolerant Distributed State Sharing. In CACM, Oct. 1996.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC